AI Weekly Malaysia

Lars Faye argues that AI coding tools create a paradox: they require expert-level judgment to use responsibly, yet they actively bypass the friction that cultivates that expertise in the first place. Developers who entered the field alongside LLMs face a 'skilled orchestrator paradox' where they need expertise they haven't had time to develop, while the tools they're told to use prevent them from developing it.

Why: If you're a junior or mid-level developer relying heavily on AI coding assistants, you may be accumulating confidence without comprehension—shipping code you can't fully evaluate. The article suggests you should deliberately seek out friction (debugging without AI, writing specs from scratch, understanding generated code line-by-line) rather than optimizing it away, because the 'good taste' needed to supervise AI output only comes from years of struggling through problems yourself.